Professional Documents
Culture Documents
A - Novita Hastuti Zen - 19218076 - Report
A - Novita Hastuti Zen - 19218076 - Report
2020
1. INTRODUCTION
1.1. Data description
Super Market Sales consisting of 1000 examples , 0 specialattributes, and 17 regular
attributes.
' With this dataset, it can be used as a reference in making some business
decisions. But net data has no value lost so that the decision to be taken can be
based on clean data, from the data I want to analyze which branches produce a lot of
sales and also what goods are sold a lot’.
1.2. Data Pre-processing
Data that is clean, not have “missing value”
From statistical data, I know that there are not have missing values so I don’t want to
clean because all of data is have imfact for other data, that is can see in Market Data
Analyze for know the relationship betIen the data with one another whether they
affect each other.
But, in here the data I want to see this have outliers or no, so to find outliers I use
the general attribute tools in Rapidminer
Before entering into the general attribute I use normalization so that the standard
deviation appears
In the function expression I write the function like the picture above so that the
outlier can be knowfrom all data the true have outlier is 10, n, and I get then the
treatment that I do to outliers by removing them because it can cause problems with
data cleaning. By using filter examples on Rapidminer I eliminate outliers
From the results of cleaning outliers, it can be seen that no outliers appear in the
data
2. PROBLEM STATEMENT
- Want to know about which branches produce a lot of sales
- Want to know what goods are sold a lot
So from all the problem, I can know strategic for the company to have many
income
- For the first step, I input our data that I will process
- After that, I set the “aggregation attribute” and “group by attributes” on the
parameter of “Aggregate” as following picture.
PREDICTIVE ANALYSIS
To analysis predictive, I choose Classic Decomposition and choose two variable that is COGS
and Gross Income to analyze forecast
USING MANUAL
I use manual analyze and calculation because many problem from excel and RapidMiner
Application.
1. Search the which branche products, so I calculate from excel
4. DATA ANALYSIS AND CONCLUSION
The many product sold is Fashion, that search with manually and get 178
sales from 1000 sales from 6 Categories of product, and from 3 city, the most many
product sold day in Yangon, different with Naypyitaw . The Company can increase
product in Yangon, and in Naypyitaw can it to but maybe in Naypyitaw can can
increase a discount
Most Popular per 1000 sales is Fashion, maybe the demand from fashion is very
big, because fashion have a big demand concent of female